Skip to content

A fully templated C++ implementation of general-use algorithms for robotic perception and visual servoing.

License

Notifications You must be signed in to change notification settings

wissem01chiha/robotics-perception

Repository files navigation

Robotic Perception Library

A fully templated C++ implementation of general-use algorithms for robotic perception and visual servoing.

C++ License Linux Build


Goals

RPL project has the following goals:

  • Provide optimized and flexible modules for collision prediction and detection for collaborative robotic applications.
  • Feature extraction from camera images and/or sensors data types.
  • Environment reconstruction and processing, trajectory planning in unconstrained worlds.
  • Robust control strategies based on visual input.
  • A set of fuzzy relations between solid objects, and develops, and a set of human-like control algorithms.

This is a template for modern C++ projects. What you get is:

  • Library, executable and test code separated in distinct folders
  • Use of modern CMake for building and compiling
  • External libraries installed and managed by
    • CPM Package Manager OR
    • Conan Package Manager OR
    • VCPKG Package Manager
  • Unit testing using Catch2 v2
  • General purpose libraries: JSON, spdlog, cxxopts and fmt
  • Continuous integration testing with Github Actions and pre-commit
  • Code documentation with Doxygen and Github Pages
  • Tooling: Clang-Format, Cmake-Format, Clang-tidy, Sanitizers

References

  • Real-Time Deep Learning Approach to Visual Servo Control and Grasp Detection for Autonomous Robotic Manipulation, E. G. Ribeiro, R. Q. Mendes, V. Grassi Jr, Elsevier's Robotics and Autonomous Systems, 2021

  • Training deep neuralnetworks for visual servoing, Q. Bateux, E. Marchand, J. Leitner, F. Chaumette, P. Corke, ICRA 2018 Library code goes into src/, main program code in app/ and tests go in tests/.

  • Classical and Deep Learning based Visual Servoing Systems: a Survey on State of the Art, * Z.Machkour, D.Ortiz-Arroyo, P.Durdevic*, 2021

Software Requirements

  • CMake 3.21+
  • GNU Makefile
  • Doxygen
  • Conan or VCPKG
  • MSVC 2017 (or higher), G++9 (or higher), Clang++9 (or higher)
  • Optional: Code Coverage (only on GNU|Clang): gcovr
  • Optional: Makefile, Doxygen, Conan, VCPKG

Building

First, clone this repo and do the preliminary work:

git clone --recursive https://github.com/franneck94/CppProjectTemplate
mkdir build
  • App Executable
cd build
cmake -DCMAKE_BUILD_TYPE=Release ..
cmake --build . --config Release --target main
cd app
./main
  • Unit testing
cmake -H. -Bbuild -DCMAKE_BUILD_TYPE="Debug"
cmake --build build --config Debug
cd build
ctest .
  • Documentation
cd build
cmake -DCMAKE_BUILD_TYPE=Debug ..
cmake --build . --config Debug --target docs
  • Code Coverage (Unix only)
cmake -H. -Bbuild -DCMAKE_BUILD_TYPE=Debug -DENABLE_COVERAGE=On
cmake --build build --config Debug --target coverage -j4
cd build
ctest .

For more info about CMake see here.

Testing

References

Contributing

please see the CONTRIBUTING guide for more informations